Saved in:
Bibliographic Details
Main Authors: Shechter, Mikey, Carmon, Yair
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.08805
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909925956911104
author Shechter, Mikey
Carmon, Yair
author_facet Shechter, Mikey
Carmon, Yair
contents We introduce Filter Like You Test (FLYT), an algorithm for curating large-scale vision-language datasets that learns the usefulness of each data point as a pretraining example. FLYT trains a scoring model that learns to weigh each example's features using gradient signals from downstream tasks training sets. Based on FLYT, we implement Mixing-FLYT (M-FLYT), which takes the per-example scores generated by different scoring methods as features, and learns to unify them into a single score. FLYT naturally produces a distribution over the training examples, which we leverage through Soft Cap Sampling (SCS), a strategy for obtaining a filtered pretraining dataset from per-example probabilities that samples examples while preventing over-representation through a repetition penalty. Using these methods, we achieve 40.1% ImageNet zero-shot accuracy on the DataComp medium scale filtering benchmark, a 2% absolute accuracy increase over all previous results and a 5.5% increase over results that - like us - use only public resources. Our approach also yields 37.7\% on the average of 38 DataComp evaluation tasks, outperforming previous public-resource approaches by 0.4\%.
format Preprint
id arxiv_https___arxiv_org_abs_2503_08805
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Filter Like You Test: Data-Driven Data Filtering for CLIP Pretraining
Shechter, Mikey
Carmon, Yair
Computer Vision and Pattern Recognition
Machine Learning
We introduce Filter Like You Test (FLYT), an algorithm for curating large-scale vision-language datasets that learns the usefulness of each data point as a pretraining example. FLYT trains a scoring model that learns to weigh each example's features using gradient signals from downstream tasks training sets. Based on FLYT, we implement Mixing-FLYT (M-FLYT), which takes the per-example scores generated by different scoring methods as features, and learns to unify them into a single score. FLYT naturally produces a distribution over the training examples, which we leverage through Soft Cap Sampling (SCS), a strategy for obtaining a filtered pretraining dataset from per-example probabilities that samples examples while preventing over-representation through a repetition penalty. Using these methods, we achieve 40.1% ImageNet zero-shot accuracy on the DataComp medium scale filtering benchmark, a 2% absolute accuracy increase over all previous results and a 5.5% increase over results that - like us - use only public resources. Our approach also yields 37.7\% on the average of 38 DataComp evaluation tasks, outperforming previous public-resource approaches by 0.4\%.
title Filter Like You Test: Data-Driven Data Filtering for CLIP Pretraining
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2503.08805